Who gets the say in the coalition agreement? Local programs and municipal coalition agreements

Nearly all research into coalition negotiations focuses on the national level. This study applies existing insights about national coalition negotiations to the local context in the Netherlands. The central question is: to what extent and under what conditions do parties negotiating to form a coalition government get policy and office pay-offs? We examine two possible explanations of policy and office pay-offs. Firstly, we examine the role that party size plays in negotiations. We know since Gamson (1961) that this play a major role in the division of offices. The broader negotiation literature also suggests that it plays an important role in determining negotiation outcomes on policy issues (Däubler and Debus, 2009; Ferland, 2016; Golder and Stramski, 2010; Indriđason, 2011; Mansergh, 1999; Martin and Vanberg, 2014; Warwick, 2001). We expect that larger parties will see more of their positions and priorities in the local council agreement than smaller parties and get more seats in the local executive. After all, larger parties have more bargaining power Secondly, we look at the role that preference intensity plays in negotiations. The existing research suggests that parties are more likely to get what they want if they care greatly about the issue and their coalition parties do not (Debus, 2008; Matthieß, 2019; Schermann and Ennser-Jedenastik, 2014a; Thomson, 2001; Zubek and Klüver, 2015; Falcó-Gimeno, 2014; Klüver and Bäck, 2019a: 2003). We therefore expect that parties are more likely to fulfil promises the higher the relative saliency for this party. In the analyses for this paper, we look at more than three thousand parties negotiating in more than 300 municipalities in three rounds of negotiations in Netherlands (2014, 2018 and 2022). For policy pay-offs, we look at more than 3,000 local election manifestos of local parties and branches of national parties that were in coalitions and more than 1,000 local coalition agreements. We apply an automated, quantitative "bag-of-words" approach to these texts. We estimate the policy positions of these texts using word scores and the attention they devote to issues using Latent Dirichlet Allocation. For office pay-offs, we look at the composition of the municipal executive both in terms of the number of offices parties get and which office they receive.